The majority of individuals that initially survive a myocardial infarction go on to suffer from life-threatening complications such as infarct rupture, arrhythmia, or heart failure. These sequelae depend critically on wound healing in the infarct as well as fibrosis in the remote myocardium, yet existing drugs are not designed to specifically target either of these. An ideal pharmacologic therapy would enhance extracellular matrix in the infarct while reducing fibrosis in the remote myocardium, but achieving such spatial control of extracellular matrix has eluded the field. We propose that these distinct local responses to globally applied drugs can be achieved by leveraging environment-dependent signaling mechanisms, uniquely identified with systems pharmacology models. We first integrate drug-target interaction databases together with a novel large-scale model of the cardiac signaling network to identify drugs or drug pairs that cause context-dependent regulation of extracellular matrix with cultured cardiac fibroblasts. We then integrate the signaling network model with agent-based models of inflammation and myocardial infarct healing. This integrated model is applied to perform multi-scale virtual drug screening to predict and then experimentally validate drugs that enhance extracellular matrix in the infarct while reducing fibrosis in the remote myocardium. Overall, this study will provide an innovative framework for multi-scale systems pharmacology of cardiac fibrosis, identify new therapeutic approaches for infarct healing, and provide proof-of-concept for spatial control of drug responses by leveraging context-dependent cell signaling.